Public SourceMage Discussion List

[Vlad Glagolev <stealth AT sourcemage.org>]

> Here's my take on the issue: revert the reversion, i.e. go back to gtk+2
> 2.14.

Okay. But that's your choice. As I already said I don't wanna have lots of 
semi-b0rked spells
in local grimoire or any other grimoire, having such issues. I stop to 
contribute to SMGL until
the problem is solved. If you're still on 2.12 or don't think about any users 
except yourself --
I just can't help with it, and I won't follow this way. I thought SMGL 
project follow the
bugless way, as I can see now -- it doesn't and I was wrong. Because of 
someone's thought "we
just need to update for update, doesn't matter if it affects the 
users/increase the number of
problems or no".

> I think this is the right solution mainly because we simply don't know
> how widespread this issue is, and if we don't update, we will never find
> out.

We _do_ and even more: we _know_. And we have a lot of info about it. Yes, 
not from SMGL users
cause I said: I just prevented the SMGL users from the problem. There are a 
lot of video chips
(mainly on laptops) affected. Xorg devs proved, that's not distro-specific. 
And even Jaka
understood: that's a _code_ problem, not any stupid config or video-card 
configuration or
something similiar.

I repeat: it's a wide-spread problem, and it's _KNOWN_:
https://bugs.launchpad.net/gtk/+bug/245383

Even more: every few hours/days the number of bugreports is increased.
And I have _THE SAME_ problem. I'm not ubuntu user (yes, really; at least 
ohloh prooves that).

Isn't the proof from gtk+ developers, xorg developers and linux users not 
enough (as we already
talked, it's not distro-specific) for you? If so, I can't talk about such 
proof anymore.

> It was a violation of our established processes to revert it without
> discussion, as it was a perfectly fine working spell.

And now you wanna revert it after the discussion. Huh, funny. So as I see, it 
wouldn't change
anything.

> We have a lot of apps and libs in the grimoire with bugs that affect one or 
> more users.

And we shouldn't keep those bugs silently if we wanna keep the community of 
users (yes, users)
increasing.

> It's simply not feasible to keep stuff at older versions until that
> isn't the case anymore, or debian stable will be more up-to-date than we
> are.

There's no any spell strictly requiring gtk+ >= 2.14 (and won't be for a 
pretty long preiod of
time, if you read/researched gtk+ roadmap). Interesting, but even ArchLinux 
devs keep gtk+
2.12. They also have glib 2.18, but only in "Testing". And the funniest thing 
is they don't
have gtk+ 2.14 _even in "Testing"_. It won't hurt the system, and the system 
won't be out of
date for a grace period of time.

> Depending on the feedback we get once this is in test, we may or may not
> revert the update in the next stable-rc, but for test grimoire, I
> consider this not big enough an issue to force this downgrade.

That's not qualified downgrade. The last release from 2.12 branch was out few 
days ago.

> We can even attach the
> old spells to the bug report for those who don't read sm-discuss.

Feel free to do so.

I repeat one more time: that's not distro-development process. Where a user 
phucks his brain
everytime he wants to get his system _just working_. We should improve the 
quality of the end-
user product, imho, not downgrading it. If a user wants to get pretty stable 
and recently
updated system: he must cast test grimoire (yes, sometimes we need the recent 
software), create
local grimoire, create gtk+ spell there, create gail spell there (yes, he 
must know how to do
that too), play with scribe. It's even more sex than with 
xorg-modular.tar.bz2.

Most of the devs (and some cool users) play and work with test grimoire.

Again and again, quoting sandalle: test grimoire isn't for testing upstream 
bugs. I don't
wanna see the grimoire as a collection of broken stuff. And I also don't 
wanna track all those
spells having issues just before tagging to next -rc.

I can't switch to stable, too. That makes me useless as developer. So I stop 
the activity.
Sorry for incomprehension.

-- 
Dont wait to die to find paradise...
--
Cheerz,
Vlad "Enqlave" Glagolev

Attachment: pgpA5zZaGlL9S.pgp
Description: PGP signature